Evaluating cardiovascular event reduction with

ezetimibe as an adjunct to simvastatin in 18,144

patients after acute coronary syndromes: Final baseline
characteristics of the IMPROVE-IT study population
Michael A. Blazing, MD, a Robert P. Giugliano, MD, SM, b Christopher P. Cannon, MD, b Thomas A. Musliner, MD, c
Andrew M. Tershakovec, MD, MPH, c Jennifer A. White, MS, a Craig Reist, PhD, a Amy McCagg, BS, b
Eugene Braunwald, MD, b and Robert M. Califf, MD a Durham, NC; Boston, MA; and Rahway, NJ

Background The IMProved Reduction of Outcomes: Vytorin Efficacy International Trial (IMPROVE-IT) is evaluating the
potential benefit for reduction in major cardiovascular (CV) events from the addition of ezetimibe versus placebo to 40 mg/d of
simvastatin therapy in patients who present with acute coronary syndromes and have low-density lipoprotein cholesterol (LDL-C)
≤125 mg/dL.
Methods The primary composite end point is CV death, nonfatal myocardial infarction (MI), nonfatal stroke,
rehospitalization for unstable angina (UA), and coronary revascularization (≥30 days postrandomization). The simvastatin
monotherapy arm’s LDL-C target is b70 mg/dL. Ezetimibe was assumed to further lower LDL-C by 15 mg/dL and produce an
estimated ~8% to 9% treatment effect. The targeted number of events is 5,250.
Results We enrolled 18,144 patients with either ST-segment elevation MI (STEMI, n = 5,192) or UA/non–ST-segment
elevation MI (UA/NSTEMI, n = 12,952) from October 2005 to July 2010. Western Europe (40%) and North America (38%)
were the leading enrolling regions. The STEMI cohort was younger and had a higher percentage of patients naive to lipid-
lowering treatment compared with the UA/NSTEMI cohort. The UA/NSTEMI group had a higher prevalence of diabetes,
hypertension, and prior MI. Median LDL-C at entry was 100 mg/dL for STEMI and 93 mg/dL for UA/NSTEMI patients.
Conclusions This trial is evaluating LDL-C lowering beyond previously targeted LDL-C levels. The results depend on
achieving the desired separation of LDL-C with ezetimibe and on the assumption that ezetimibe’s lowering of LDL-C will have
similar event reduction efficacy as the LDL-C lowering from a statin. The results could affect future therapies and guidelines.
(Am Heart J 2014;168:205-212.e1.)

Individual trials and meta-analyses have demonstrated fer protein inhibitors also alter serum lipid profiles in
that more aggressive treatment with statins, using either directions that have been presumed beneficial. When
more potent drugs or higher doses, results in decreased evaluated as adjuncts to aggressive statin therapy in trials
low-density lipoprotein cholesterol (LDL-C) levels and a designed to assess effects on CV outcomes, these
further lowering of cardiovascular (CV) event rates nonstatin agents produced their expected complementa-
when compared with less potent or lower dose statin ry or additive lipid effects but failed to achieve their
therapies. 1-6 Niacin, fibrates, and cholesteryl ester trans- predicted effects on CV event reduction. 7-9
Ezetimibe is a nonstatin agent commonly used as an
adjunctive therapy in combination with statins to further
From the aDuke Clinical Research Institute, Durham, NC, bThrombolysis in Myocardial
lower LDL-C. It inhibits the intestinal absorption of
Infarction [TIMI] Study Group, Cardiovascular Division, Brigham and Women’s Hospital, cholesterol, leading to an upregulation of LDL receptors
Boston, MA, and cMerck & Co., Rahway, NJ. in the liver, which results in lowering of LDL-C in the
Roger S. Blumenthal, MD, served as guest editor for this article.
serum. 10 This LDL-C lowering is independent and
Submitted April 18, 2014; accepted May 13, 2014.
Reprint requests: Michael A. Blazing, MD, Duke University Medical Center, Box 3126,
additive to that of a statin. 11 A large, pooled analysis
Durham, NC 27710. found that adding ezetimibe to ongoing statin therapy
E-mail: michael.blazing@dm.duke.edu resulted in an average 23.4% further reduction in LDL-C
0002-8703
relative to the LDL-C level attained with statin monother-
© 2014, The Authors. Published by Mosby, Inc. This is an open access article under the CC
BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/). apy before addition of ezetimibe. 12 Furthermore, unlike
http://dx.doi.org/10.1016/j.ahj.2014.05.004 niacin, fibrates, or cholesteryl ester transfer protein
 American Heart Journal
206 Blazing et al August 2014

Figure 1

IMPROVE-IT study design. ASA, aspirin. Adapted from Cannon et al, Am Heart J 2008;156:826-832.

inhibitors when used as an adjunctive therapy, ezetimibe sufficient risk as defined in the protocol (outlined below)
has been shown to augment the effect of statins on and who had an initial LDL-C of ≤125 mg/dL if lipid-
lowering of high-sensitivity C-reactive protein. 12 Clinical- lowering naive or b100 mg/dL if on a prior prescription
ly, the combination of ezetimibe with simvastatin lipid-lowering therapy identified as no more potent
compared with placebo has been shown to reduce first than simvastatin 40 mg/d. All patients received simvastatin
clinical cardiovascular event in patients with aortic at a starting daily dose of 40 mg and either placebo or 10 mg
stenosis who had no known coronary disease 13,14 and of ezetimibe added to the baseline simvastatin therapy. The
to reduce a primary composite end point of first major LDL-C entry limitations were set to achieve a mean LDL-C of
atherosclerotic event (nonfatal myocardial infarction [MI] b70 mg/dL in the simvastatin/placebo cohort, which was
or coronary death, nonhemorrhagic stroke, or any arterial the optional recommended target set in the last update of
revascularization procedure) in persons with moderate the Adult Treatment Panel III guidelines. 17 The
chronic kidney disease. 15 primary end point is CV death, nonfatal MI, nonfatal
The IMProved Reduction of Outcomes: Vytorin Efficacy stroke, rehospitalization for unstable angina (UA), and
International Trial (IMPROVE-IT) was designed to deter- coronary revascularization (occurring at least 30 days
mine whether adding ezetimibe to simvastatin in patients after randomization).
presenting with acute coronary syndromes (ACS) adds
clinical benefit by further reducing major CV events Study population
compared with simvastatin monotherapy. 16 The trial Initially, the trial enrolled high-risk patients with ST-
design and characteristics of the first 10,000 enrolled segment elevation MI (STEMI), non–ST-segment elevation
patients have been described previously. 16 This article MI (NSTEMI), or documented UA who had stable
describes the baseline characteristics of the complete hemodynamics, arrhythmias, and ischemic symptoms
cohort enrolled in IMPROVE-IT and discusses the implica- and did not require medications known at the time of
tions of nonstatin-mediated LDL-C reduction and CV risk protocol development in 2004 to interact with simvastat-
reduction relating to recent clinical outcomes data. in (see online Appendix). The protocol-specified high-
risk characteristics for STEMI were anterior MI or age
≥50 years. The high-risk characteristics for UA/NSTEMI
Methods were age ≥50 years and one of the following: ST-segment
Study design and objectives changes of at least 1 mm, positive cardiac biomarkers,
The design of IMPROVE-IT (ClinicalTrials.gov, diabetes mellitus, a history of previous MI, a history of
NCT00202878) is outlined in Figure 1. 16 The study enrolled coronary artery bypass grafting at least 3 years earlier, or
patients within 10 days of ACS hospitalization who had demonstration of at least 2 major coronary arteries with
American Heart Journal
Volume 168, Number 2
Blazing et al 207

≥50% luminal narrowing. Patients enrolled in the Early Table I. Protocol amendments
Glycoprotein IIb/IIIa Inhibition in Non-ST-Segment Ele- Amendment
vation ACS (EARLY ACS) trial 18 were eligible to be number Date Major focus
enrolled as well.
Enrollment of STEMI patients was phased out beginning 1 April Changes to inclusion and
in September 2007 with a second protocol amendment. 2007 exclusion criteria 16
2 September Capped enrollment of STEMI 16
The STEMI enrollment was limited to minimize any
2007
potential effects of lower long-term risk of this population 3 March Rationale and plan for
on trial duration (Figure 2). 2008 sample size readjustment 16
4 and 5 June 2011 Response to restrictions on simvastatin
80 mg/d use and a second interim
Treatment protocol analysis at 75% of events

The original protocol outlined monitoring of patients’

LDL-C levels during the trial and called for their dose of was continued rather than reducing it to the FDA-
simvastatin to be raised to 80 mg/d if 2 successive LDL-C recommended limitation of 20 mg/d for patients taking
values exceeded 79 mg/dL. Minimum follow-up was amlodipine or ranolazine based on safety data from this
specified to be 2.5 years after the last enrolled patient. trial. An independent events committee unaware of
In September 2008, an association between ezetimibe treatment assignment adjudicated and classified all mus-
and increased risk for cancer was reported in a trial cle-related events identified on adverse event and serious
evaluating the use of simvastatin with ezetimibe on the adverse event reports using prespecifed criteria described
progress of aortic stenosis (Simvastatin and Ezetimibe in in the protocol. The data from this committee, stratified by
Aortic Stenosis [SEAS] 14). To address this concern, a treatment assignment, were evaluated in the safety reviews
specific cancer-related clinical events committee was of the DSMB. Patients stable on simvastatin 80 mg/d for ≥1
formed to provide more detailed cancer-specific data year were left at that dose per the FDA advisory. These
regarding all prevalent and incident cancer cases within protocol changes and a second interim analysis at 75% of
IMPROVE-IT to the independent data and safety monitor- accumulated target events resulted in a combined fourth
ing board (DSMB) monitoring the trial. The data from this and fifth amendment to the protocol.
clinical events committee were reviewed by the trial
DSMB a total of 7 times between 2008 and the end of the
trial with no new safety concerns. Statistical design and analysis
Overall, the protocol has undergone 5 amendments Key assumptions about sample size and targeted
(Table I). The first 3 of these have been previously expectations about events included an estimated 1%
described and (1) were related to refinements of the clinical benefit for every 1.6 mg/dL difference in LDL-C, 1
inclusion/exclusion criteria, (2) set the cap on enrollment a 15 mg/dL incremental reduction in LDL-C with
of STEMI patients, and (3) described the rationale of ezetimibe treatment, and a 25% discounting of treatment
increasing sample size from an original value of 10,000 to effect in the first 6 months. 16 These assumptions yielded
the final value of approximately 18,000. 16 Modeling was an estimated treatment effect of 9.375% for the primary
used to arrive at this number of patients, which was found end point. An independent lipid monitoring committee
to be optimal for preserving study power while minimizing reviewed on-treatment LDL-C values in the context of the
potential issues concerning duration of study follow-up and study power assumptions. The committee did not
enrollment. Since 2009, the protocol has undergone 2 recommend any changes in sample size or trial duration.
additional amendments to accommodate changes to the There were 3 interim analyses for efficacy during the
simvastatin product label and to add a second interim trial. The original protocol-specified interim analysis at
analysis. In June 2011, a Food and Drug Administration 50% of targeted end points was performed in March 2010
(FDA) safety communication restricted the use of 80 mg/d and recommended continuation of the trial. The second
of simvastatin to patients who had been on that dose and DSMB interim analysis at 75% of targeted end points
stable for 12 months, contraindicated any new starts of included in the fifth protocol amendment was performed
simvastatin 80 mg/d, and identified amlodipine and in March 2012. At that time, the DSMB recommended
ranolazine as new drugs requiring simvastatin dosage continuation of the trial but requested a third analysis for
limitations. 19 This communication led to the discontinua- efficacy, which took place in March 2013 with the same
tion of protocol-directed use of 80 mg/d of simvastatin for recommendations as after the 2012 analysis. Stopping
(1) patients who had an LDL-C of N79 mg/dL on simvastatin guidelines for “overwhelming efficacy” were in place
40 mg/d, (2) patients who had been on simvastatin 80 mg/d from the outset and were based on findings for the
for b1 year, or (3) patients who needed to continue either primary end point in conjunction with a directionally
amlodipine or ranolazine. In these 3 groups, the simvastatin consistent reduction in total mortality. There were no
dose was restricted or reduced to 40 mg/d. The 40-mg dose stopping rules for futility provided to the DSMB for any of
 American Heart Journal
208 Blazing et al August 2014

Figure 2

Enrollment into IMPROVE-IT.

the interim analyses. Using East Software (Version 5.3; 64 years) and had lower prevalence of hypertension,
Cytel, Cambridge, MA), the nominal alpha level for the diabetes mellitus, and previous MI. Substantially more
final primary end point analysis was adjusted to 0.0394 to STEMI patients were not on lipid-lowering therapy at
account for the 3 interim analyses. entry (83%) compared with the UA/NSTEMI cohort (60%).
The trial was initially funded by Schering-Plough and Forty-nine percent of the STEMI patients had anterior
Merck and then by Merck & Co alone after their merger in infarction as a high-risk feature (data not shown). Diabetes
November 2009. (30%) was the most common high-risk feature qualifying
patients with UA/NSTEMI, followed by previous MI (26%).
Baseline laboratory findings are also shown in Table II.
The trial enrolled patients in the target LDL-C range, and
Results the median high-density lipoprotein cholesterol (HDL-C)
Enrollment began in October 2005 and completed in level was 40 mg/dL. The median LDL-C level was lower in
July 2010 (Figure 2). The shape of the overall enrollment the UA/NSTEMI cohort compared with the STEMI cohort
pace reflects the change in sample size to 18,000 patients, (93 mg/dL vs 100 mg/dL, P b .001). Renal function was
made 2.5 years into enrollment and implemented in well preserved, with a median serum creatinine of
March 2008, and the closeout of STEMI enrollment that 1.0 mg/dL, and 75% of patients had a calculated
started in September 2007. Median enrollment was creatinine clearance of ≥65 mL/min.
reached in October 2007. The target was met with The distributions of baseline lipid levels are shown in
18,144 patients enrolled at 1,148 sites in 39 countries. Figure 3. They demonstrate a narrow range of LDL-C
The 2 highest enrolling regions were Western Europe levels, as anticipated from the eligibility criteria. Approx-
(40%) and North America (38%). The prespecified imately two-thirds of the patients enrolled were naive to
minimum trial duration was reached in January 2013. lipid-lowering therapy. The lipid-lowering–naive patients
The targeted number of 5,250 events is projected to be had a median LDL-C level that was higher (104 mg/dL)
reached in late spring of 2014. compared with the median LDL-C (80 mg/dL) of those on
Baseline demographic and medical history characteris- prior lipid-lowering therapy (data not shown).
tics are shown in Table II. The median age at
randomization was 63 years, with an interquartile range
of 56 to 71 years. Most of the patients are male (76%). At
randomization, 27% of patients had diabetes, and 21% had Discussion
experienced a previous MI. The initial event was STEMI in The IMPROVE-IT trial enrolled patients with selected
5,192 (29%), NSTEMI in 8,567 (47%), and UA in 4,385 high-risk criteria after stabilization of a qualifying ACS
(24%) patients. Compared with the UA/NSTEMI cohort, event who presented with an LDL-C level ≤125 mg/dL.
the STEMI cohort was younger (median of 60 years vs The patients are from countries with different clinical
American Heart Journal
Volume 168, Number 2
Blazing et al 209

Table II. Baseline characteristics

STEMI UA/NSTEMI All subjects

Age at randomization (y)

n 5192 12952 18144
Median (IQR) 60 (54-68) 64 (57-72) 63 (56-71)
Male 4155 (80.0%) 9573 (74%) 13728 (76%)
Race
American Indian or Alaskan Native 14 (0.3%) 38 (0.3%) 52/18141 (0.3%)
Asian 157 (3%) 616 (5%) 773/18141 (4%)
Black 106 (2%) 392 (3%) 498/18141 (3%)
Spanish descent 270 (5%) 538 (4%) 808/18141 (5%)
Native Hawaiian or Pacific Islander 9 (0.2%) 10 (0.1%) 19/18141 (0.1%)
Caucasian 4458 (86%) 10745 (83%) 15203/18141 (84%)
Other 174 (3%) 597 (5%) 771/18141 (4%)
Collection prohibited 3 (0.1%) 14 (0.1%) 17/18141 (0.1%)
Region
United States 1928 (37%) 3938 (30%) 5866 (32%)
Canada 434 (8%) 673 (5%) 1107 (6%)
Western Europe 2071 (40%) 5203 (40%) 7274 (40%)
Eastern Europe 224 (4%) 1192 (9%) 1416 (8%)
Malaysia/Singapore/Hong Kong 89(2%) 527 (4%) 616 (3%)
South America 393 (8%) 1192 (9%) 1585 (9%)
Australia/New Zealand 53 (1%) 227 (2%) 280 (2%)
Hypertension 2498 (48%) 8640 (67%) 11138 (61%)
Diabetes 1005 (19%) 3911 (30%) 4916 (27%)
Previous MI 486 (9%) 3320 (26%) 3806 (21%)
Lipid-lowering naive 4321 (83%) 7742 (60%) 12063 (67%)
LDL-C (mg/dL)
n 5156 12843 17999
Median (IQR) 100 (85-113) 93 (77-108) 95 (79-110)
HDL-C (mg/dL)
n 5142 12717 17859
Median (IQR) 40 (33-49) 40 (33-49) 40 (33-49)
Triglycerides (mg/dL)
n 5142 12762 17904
Median (IQR) 114 (79-165) 124 (88-175) 120 (85-172)
Creatinine (mg/dL)
n 5159 12872 18031
Median (IQR) 1.0 (0.8-1.2) 1.0 (0.8-1.2) 1.0 (0.8-1.2)
Creatinine clearance (mL/min)
n 5136 12836 17972
Median (IQR) 87 (69-110) 83 (64-105) 84 (65-106)

Abbreviation: IQR, interquartile range.

practice patterns and varying social and economic average) the Adult Treatment Panel III/European guide-
backgrounds, which should make the findings from line–based goal of b70 mg/dL. 17,20
the trial widely applicable. As expected, the STEMI and The outcome depends on ezetimibe producing the
UA/NSTEMI ACS populations differ with respect to age targeted LDL-C difference between treatment arms and
and risk factors. The median lipid values for the this difference having proportionally the same effect on
study population were within expected ranges (LDL-C outcomes as would be expected with a statin. Using
95 mg/dL, HDL-C 40 mg/dL, triglycerides 120 mg/dL) ezetimibe in addition to a statin produces an average
given the LDL-C entry criteria. The lower LDL-C and incremental LDL-C reduction of 23% to 24% relative to on-
higher triglyceride levels in the UA/NSTEMI population statin LDL-C, an effect similar to that achieved by an 8-fold
possibly result from differences in previous statin use and increase in simvastatin statin dose (ie, 3 dose doublings
in prevalence of diabetes between the 2 populations. The from 10 to 80 mg). 21 Thus, for the IMPROVE-IT population,
results should provide an evaluation of whether ezeti- where simvastatin monotherapy was designed to achieve a
mibe, with its modest estimated incremental reduction in median LDL-C of slightly b70 mg/dL, the addition of
LDL-C of 15 mg/dL in this trial, produces further ezetimibe should produce the targeted 15 mg/dL differ-
reduction in CV events in comparison with a simvastatin ence in LDL-C between the treatment groups. This delta
monotherapy control population targeted to achieve (on for LDL-C lies at the lower limit of the range of other more
 American Heart Journal
210 Blazing et al August 2014

Figure 3 Table III. Comparison of more intensive statin therapy trials

LDL-C Delta
Trial n LDL-C control aggressive LDL-C

A to Z 3 4496 77 mg/dL 63 mg/dL 14 mg/dL

phase Z
PROVE IT4 4162 95 mg/dL 62 mg/dL 33 mg/dL
TNT 6 10001 100 mg/dL 77 mg/dL 23 mg/dL
IDEAL 5 8888 104 mg/dL 81 mg/dL 23 mg/dL
IMPROVE-IT 18154 b70 mg/dL b55 mg/dL ~15 mg/dL

Abbreviations: A to Z, Aggrastat to Zocor; TNT, Treating to New Targets; IDEAL,

Incremental Decrease in Endpoints through Aggressive Lipid Lowering.

ence of 15 mg/dL in LDL-C due to ezetimibe is estimated to

fall in the range of 8% to 9%.
A variety of factors could potentially influence the
treatment effect of ezetimibe in IMPROVE-IT. First is the
assumption that the relationship between LDL-C lowering
and CV event lowering with statins will be the same for
ezetimibe. The 2 drugs work through final pathways that
lead to upregulation of LDL receptors that mediate a
complementary reduction in serum LDL-C, and the
combination of ezetimibe with ongoing statin augments
lowering of C-reactive protein 11,12; however, the incre-
mental efficacy of adding ezetimibe on biomarkers such
as carotid intimal thickening and vascular reactivity has
been inconsistent, with studies showing beneficial,
neutral, and possibly detrimental effects of the drug
despite its consistent lowering of LDL-C in those
studies. 22-24 Second, there is question whether the
relationship between LDL-C lowering and reduction in
CV risk (based on meta-analyses of statin trials 2) will
remain linear over the entire comparison range of LDL-C
being evaluated in IMPROVE-IT. 25 In a post hoc analysis
of a similar post-ACS population enrolled in the Pravastatin
or Atorvastatin Evaluation and Infection Therapy (PROVE IT)
trial (n = 4,162), 4 an examination of statin-naive patients (n =
2,986) stratified by quartiles of entry LDL-C found the benefit
of intensive therapy progressively declined as untreated
baseline entry LDL-C decreased. 25 This study suggests that
Distribution of LDL-C A, HDL-C B, and triglycerides C at entry. Q1,
the lower LDL-C entry criteria in IMPROVE-IT compared
Q3, first and third quartiles.
with PROVE IT (LDL-C limited to ≤125 mg/dL vs no limit,
respectively) could make the IMPROVE-IT results more
susceptible to any effects of flattening of the relationship
between efficacy and LDL-C that may occur at lower LDL-C
levels. Interpretation of a separate analysis from PROVE IT,
intensive statin therapy trials 3-6 (Table III). The IMPROVE- in which cohorts of patients who achieved LDL-C of ≤60
IT trial is larger than these previous trials, with longer mg/dL were found to have fewer major CV events than
follow-up (~6 years vs 2-5 years), and will have substantially cohorts not reaching these levels, 26 suggests that the
more primary end points (5,250) than all of these other ezetimibe arm of IMPROVE-IT (with its projected mean
studies combined. Based on data derived from these and achieved LDL-C b55 mg/dL) could have fairly low event
other statin trials included in the Cholesterol Treatment rates. These lower rates could help preserve a linear
Trialists’ meta-analysis 2 (Figure 4), the incremental out- relationship between LDL-C and efficacy at lower levels of
comes reduction attributable to the between-group differ- achieved LDL-C. Although it is unclear which marker—
American Heart Journal
Volume 168, Number 2
Blazing et al 211

Figure 4 The implications of the results from IMPROVE-IT go

beyond the specific circumstances of the trial. Although
some may question the clinical utility of small treatment
effects at low achieved LDL-C levels, the practical value of a
positive trial will serve as a definitive proof of concept that
LDL-C lowering with compounds other than statins can
affect outcomes. The trial may provide support for driving
LDL-C to levels below those previously recommended and
will raise the issue of whether other molecules—such as
PCSK9 antibodies—could have similar proportional benefit
as an adjunct to statins. A positive result will also support
the concept that ezetimibe can be effective in patients who
are unable to tolerate high-dose statins, those who may
better tolerate a combination of low-dose statin plus
ezetimibe, and those who do not achieve adequate
LDL-C lowering despite high-dose statin use. A positive
result could also lead to an early re-evaluation of the new
American College of Cardiology/American Heart Associ-
ation cholesterol guidelines that endorse statins as the
only recommended drugs for treating cholesterol-related
CV risk. 28 If IMPROVE-IT does not meet its primary end
Projected positioning of IMPROVE-IT in comparison with the relationship point, subsequent analyses evaluating the trial popula-
of LDL-C difference and proportional event reduction for intensive versus tion, drug discontinuations, and assumptions made
less-intensive statin therapy trials. Adapted from a slide available on the regarding treatment effects will provide insight into
SHARP trial website (http://www.sharpinfo.org/) and presented at the possible trial design issues or mechanistic revelations
FDA Endocrinologic and Metabolic Drugs Advisory Committee Meeting that could certainly affect beliefs about the “lower is
in November 2011 (http://www.fda.gov/downloads/Advisory better” hypothesis. Regardless of the outcome with
Committees/CommitteesMeetingMaterials/Drugs/Endocrinologicand respect to ezetimibe, a trial with 18,144 patients that
MetabolicDrugsAdvisoryCommittee/UCM279296.pdf) (From www. accumulated 5,250 events and had a mean follow-up time
fda.gov: “Unless otherwise noted, the contents of the FDA website approaching 6 years will add to our understanding of
(www.fda.gov)—both text and graphics—are not copyrighted. They are post-ACS and long-term care of atherosclerotic CV
in the public domain and may be republished, reprinted and otherwise disease in current practice.
used freely by anyone without the need to obtain permission from FDA.
Credit to the U.S. Food and Drug Administration as the source is
appreciated but not required.”).
Disclosures
Dr Blazing has served as an advisory board member for
statin-naive LDL-C at entry or achieved LDL-C—is more Merck, has consulted for AstraZeneca and Novartis, and
important or a better predictor of outcomes, it is certain has received grant support from Merck (to the Duke
that the data from IMPROVE-IT will add further insight into Clinical Research Institute). Dr Giugliano has received
the markers’ relative importance and that the outcome of grant support from Amgen, Daiichi-Sankyo, and Merck (to
IMPROVE-IT could be affected by either. Finally, any the Thrombolysis in Myocardial Infarction [TIMI] Study
relationship between the measured LDL-C difference and Group) and honoraria from Amgen, Beckman-Coulter,
measured treatment effect will have to account for possible Bristol-Myers Squibb, Daiichi-Sankyo, Janssen, Lexicon,
effects of potential trial-related issues such as withdrawn Merck, Regeneron, and Sanofi. Dr Cannon has received
consent and patient drug discontinuation. For example, in research grants and support from Accumetrics, Arisaph,
IMPROVE-IT, the median trial duration of almost 6 years, AstraZeneca, Boehringer-Ingelheim, CSL Behring, Essen-
new FDA regulations 19 that required changes in the dosing tialis, GlaxoSmithKline, Janssen, Merck, Regeneron,
of simvastatin during the trial, and the 2 negative episodes Sanofi, and Takeda and has served on an advisory board
of publicity surrounding other ezetimibe trials that closed for Bristol-Myers Squibb, Lipimedix, and Pfizer.
during IMPROVE-IT (re. drug efficacy concerns in the Drs Musliner and Tershakovec are employees and
Ezetimibe and Simvastatin in Hypercholesterolemia En- stockholders of Merck. Dr Braunwald has received
hances Atherosclerosis Regression [ENHANCE] trial 27 and grant support from AstraZeneca, Johnson & Johnson,
drug safety concerns of a cancer risk in SEAS 14) are all Beckman-Coulter, Daiichi-Sankyo, Merck, Roche Diag-
factors that could produce issues with patient drug nostics, Sanofi-Aventis, GlaxoSmithKline, and Bristol-
discontinuation or withdrawal of consent. Myers Squibb (to the TIMI Study Group) and honoraria
 American Heart Journal
212 Blazing et al August 2014

from Amorcyte, The Medicines Company, Medscape, International Trial): comparison of ezetimbe/simvastatin versus
Bayer, Daiichi-Sankyo, Menarini International, and Sanofi- simvastatin monotherapy on cardiovascular outcomes in patients with
Aventis. Dr Califf’s disclosures are available at https:// acute coronary syndromes. Am Heart J 2008;156:826-32.
17. Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical
www.dcri.org/about-us/conflict-of-interest. The other au-
trials for the National Cholesterol Education Program Adult Treatment
thors have no disclosures.
Panel III guidelines. Circulation 2004;110:227-39.
18. Giugliano RP, Newby LK, Harrington RA, et al. The early glycoprotein
IIb/IIIa inhibition in non-ST-segment elevation acute coronary
References syndrome (EARLY ACS) trial: a randomized placebo-controlled trial
1. Cannon CP, Steinberg BA, Murphy SA, et al. Meta-analysis of evaluating the clinical benefits of early front-loaded eptifibatide in the
cardiovascular outcomes trials comparing intensive versus moderate treatment of patients with non-ST-segment elevation acute coronary
statin therapy. J Am Coll Cardiol 2006;48:435-45. syndrome—study design and rational. Am Heart J 2005;149:
2. Baigent C, Blackwell L, Emberson J, et al. Efficacy and safety of more 994-1002.
intensive lowering of LDL cholesterol: a meta-analysis of data from 19. US Food and Drug Administration. FDA Drug Safety Communication:
170,000 participants in 26 randomised trials. Lancet 2010;376: new restrictions, contraindications, and dose limitations for Zocor
1670-81. (simvastatin) to reduce the risk of muscle injury. 2011. [http://www.
3. de Lemos JA, Blazing MA, Wiviott SD, et al. Early intensive vs a delayed fda.gov/drugs/drugsafety/ucm256581.htm. Last accessed
conservative simvastatin strategy in patients with acute coronary December 13, 2013].
syndromes: phase Z of the A to Z trial. JAMA 2004;292:1307-16. 20. Graham I, Atar D, Borch-Johnsen K, et al. European guidelines on
4. Cannon CP, Braunwald E, McCabe CH, et al. Intensive versus cardiovascular disease prevention in clinical practice: executive
moderate lipid lowering with statins after acute coronary syndromes. summary: Fourth Joint Task Force of the European Society of
N Engl J Med 2004;350:1495-504. Cardiology and Other Societies on Cardiovascular Disease
5. Pedersen TR, Faergeman O, Kastelein JJ, et al. High-dose atorvastatin Prevention in Clinical Practice (constituted by representatives of nine
vs usual-dose simvastatin for secondary prevention after myocardial societies and by invited experts). Eur Heart J 2007;28:2375-414.
infarction: the IDEAL study: a randomized controlled trial. JAMA 21. Davidson MH, McGarry T, Bettis R, et al. Ezetimibe coadministered
2005;294:2437-45. with simvastatin in patients with primary hypercholesterolemia. J Am
6. LaRosa JC, Grundy SM, Waters DD, et al. Intensive lipid lowering with Coll Cardiol 2002;40:2125-34.
atorvastatin in patients with stable coronary disease. N Engl J Med 22. Toth PP, Catapano A, Tomassini JE, et al. Update on the efficacy and
2005;352:1425-35. safety of combination ezetimibe plus statin therapy. Clin Lipidol
7. The ACCORD Study Group. Effects of combination lipid therapy in 2010;5:655-84.
type 2 diabetes mellitus. N Engl J Med 2010;362:1563-74. 23. Fleg JL, Mete M, Howard BV, et al. Effect of statins alone versus statins
8. Schwartz GG, Olsson AG, Abt M, et al. Effects of dalcetrapib in plus ezetimibe on carotid atherosclerosis in type 2 diabetes: the
patients with a recent acute coronary syndrome. N Engl J Med SANDS (Stop Atherosclerosis in Native Diabetics Study) trial. J Am
2012;367:2089-99. Coll Cardiol 2008;52:2198-205.
9. The AIM-HIGH Investigators. Niacin in patients with low HDL 24. Villines TC, Stanek EJ, Devine PJ, et al. The ARBITER 6-HALTS Trial
cholesterol levels receiving intensive statin therapy. N Engl J Med (Arterial Biology for the Investigation of the Treatment Effects of
2011;365:2255-67. Reducing Cholesterol 6-HDL and LDL Treatment Strategies in
10. Davis HR, Veltri EP. Zetia: inhibition of Niemann-Pick C1 Like 1 Atherosclerosis): final results and the impact of medication adherence,
(NPC1L1) to reduce intestinal cholesterol absorption and treat dose, and treatment duration. J Am Coll Cardiol 2010;55:2721-6.
hyperlipidemia. J Atheroscler Thromb 2007;14:99-108. 25. Giraldez RR, Giugliano RP, Mohanavelu S, et al. Baseline low-density
11. Phan BA, Dayspring TA, Toth PP. Ezetimibe therapy: mechanism of lipoprotein cholesterol is an important predictor of the benefit of
action and clinical update. Vasc Health Risk Manag 2012;8:415-27. intensive lipid-lowering therapy: a PROVE IT-TIMI 22 (Pravastatin or
12. Morrone D, Weintraub WS, Toth PP, et al. Lipid-altering efficacy of Atorvastatin Evaluation and Infection Therapy–Thrombolysis In
ezetimibe plus statin and statin monotherapy and identification of factors Myocardial Infarction 22) analysis. J Am Coll Cardiol 2008;52:
associated with treatment response: a pooled analysis of over 21,000 914-20.
subjects from 27 clinical trials. Atherosclerosis 2012;223:251-61. 26. Wiviott SD, Cannon CP, Morrow DA, et al. Can low-density
13. Holme I, Boman K, Brudi P, et al. Observed and predicted reduction of lipoprotein be too low? The safety and efficacy of achieving very low
ischemic cardiovascular events in the Simvastatin and Ezetimibe in low-density lipoprotein with intensive statin therapy: a PROVE IT-TIMI
Aortic Stenosis trial. Am J Cardiol 2010;105:1802-8. 22 substudy. J Am Coll Cardiol 2005;46:1411-6.
14. Rossebø AB, Pedersen TR, Boman K, et al. Intensive lipid lowering with 27. Kastelein JP, Akdim F, Stroes ES, et al. Simvastatin with or without
simvastatin and ezetimibe in aortic stenosis. N Engl J Med 2008;359: ezetimibe in familial hypercholesterolemia. N Engl J Med 2008;358:
1343-56. 1431-43.
15. Baigent C, Landray MJ, Reith C, et al. The effects of lowering LDL 28. Stone NJ, Robinson J, Lichtenstein AH, et al. ACC/AHA guideline on
cholesterol with simvastatin plus ezetimibe in patients with chronic the treatment of blood cholesterol to reduce atherosclerotic
kidney disease (Study of Heart and Renal Protection): a randomised cardiovascular risk in adults: a report of the American College of
placebo-controlled trial. Lancet 2011;377:2181-92. Cardiology/American Heart Association Task Force on Practice
16. Cannon CP, Giugliano RP, Blazing MA, et al. Rationale and design of Guidelines. J Am Coll Cardiol 2013. http://dx.doi.org/10.1016/j.
IMPROVE-IT (IMProved Reduction of Outcomes: Vytorin Efficacy jacc.2013.11.002. [Advance online publication: November 2013].
American Heart Journal
Volume 168, Number 2
Blazing et al 212.e1

Appendix. Prohibited medications at

trial inception

Medication
Probucol
Amiodarone
Cyclosporine
Fibric acid derivatives (fibrates)
Resins
Niacin N100 mg/d
Danazol
Antifungal azoles via oral and parenteral administration
(itraconazole, fluconazole, and ketoconazole)
Macrolide/ketolide antibiotics via oral and parenteral
administration (eg, clarithromycin, erythromycin, telithromycin)
Protease inhibitors
Nefazodone
Any investigational drugs
Diltiazem
Verapamil
Statins
Ezetimibe
Fusidic acid
Torcetrapib, within 1 y before screening/randomization

Short-term use of these medications, except for investigational drugs, and-lipid

lowering agents was permitted provided the study drug could be stopped during
administration and restarted after completion of the course of therapy.